Project Summary There is growing recognition that neurodegenerative brain changes in conditions such as Alzheimer's disease (AD) occur years before symptoms are currently detected using traditional cognitive assessment paradigms. These are not sufficiently sensitive to detect the subtle cognitive changes associated with the condition that occur early in the disease continuum. There is a rising consensus in the field that novel cognitive assessment paradigms are critically needed to serve as more sensitive clinical outcome measures in MCI-AD clinical trials. Capturing deficits early is necessary to move the field forward as it makes efforts to develop both prevention strategies and novel treatment approaches, which are likely to be most effective in the earliest stages of disease. In addition to the above limitations, widely used paper-and-pencil measures employed for the detection of AD-related Mild Cognitive Impairment (MCI) require a highly skilled examiner, and are staff and time intensive. Existing computerized batteries too have their limitations in that many employ insensitive measurement paradigms, and often are unavailable for diverse cultural/language groups. Our group are leaders in the field developing ?cognitive stress tests,? and have demonstrated that stressing the cognitive system by eliciting proactive semantic interference (PSI) and then, measuring recovery from PSI represents a more sensitive approach to detecting cognitive decline during the preclinical stages of AD. Moreover, vulnerability to PSI and failure to recover from PSI has been highly associated with brain changes on MRI as well as amyloid load in AD signature regions. These important findings stemming from the paper-and-pencil versions of our cognitive tests have been refined further by selecting the most sensitive indices and computerizing the instruments to reduce time and error, and increase accessibility and cross- cultural applicability. We propose to recruit 250 community-dwelling older adults to conduct a longitudinal examination of the utility of three novel computerized cognitive stress tests to detect amnestic Mild Cognitive Impairment (MCI) versus normal cognition. These novel cognitive tests, the Loewenstein-Acevedo Scales of Semantic Interference- Brief Computerized version (LASSI-BC), Miami Prospective Memory Test- Computerized version (MPMT-C), and Miami Test of Semantic Interference and Learning, Extended version (MITSI-LE) will be compared to ubiquitously used computerized cognitive measures in AD clinical trials among Hispanic and Non-Hispanic individuals. Further, we will relate performance on these novel measures to regional brain changes on MRI in AD signature regions, and ApoE genotype. We predict that among individuals with amnestic MCI, baseline and change scores on the LASSI-BC, MPMT-C, and MITSI-LE will exhibit stronger associations with increasing brain atrophy on MRI over a three-year period relative to traditional and widely used computerized neuropsychological measures that are currently available.